Causality in structural engineering: discovering new knowledge by tying induction and deduction via mapping functions and explainable artificial intelligence

Abstract

Causality is the science of cause and effect. It is through causality that explanations can be derived, theories can be formed, and new knowledge can be discovered. This paper presents a modern look into establishing causality within structural engineering systems. In this pursuit, this paper starts with a gentle introduction to causality. Then, this paper pivots to contrast commonly adopted methods for inferring causes and effects, i.e., induction (empiricism) and deduction (rationalism), and outlines how these methods continue to shape our structural engineering philosophy and, by extension, our domain. The bulk of this paper is dedicated to establishing an approach and criteria to tie principles of induction and deduction to derive causal laws (i.e., mapping functions) through explainable artificial intelligence (XAI) capable of describing new knowledge pertaining to structural engineering phenomena. The proposed approach and criteria are then examined via a case study.

Keywords

Causality / Explainable artificial intelligence / Mapping functions / Knowledge discovery / Structural engineering

References

[1] AISC. (2017). Steel Construction Manual | American Institute of Steel Construction. AISC.

[2] AltmannA, ToloşiL, SanderO, LengauerT. Permutation importance: A corrected feature importance measure. Bioinformatics, 2010, 26(10):1340-1347

[3] BabanajadSK, GandomiAH, AlaviAH. New prediction models for concrete ultimate strength under true-triaxial stress states: An evolutionary approach. Advances in Engineering Software, 2017, 110: 55-68

[4] BentzEC, VecchioFJ, CollinsMP. Simplified modified compression field theory for calculating shear strength of reinforced concrete elements. ACI Structural Journal, 2006, 103(4):614-624

[5] Bijelić, N., Lin, T., & Deierlein, G. (2019). “Classification algorithms for collapse prediction of tall buildings and regional risk estimation utilizing SCEC CyberShake simulations.” 13th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2019.

[6] BoothbyT, CloughS. Empiricist and rationalist approaches to the design of concrete structures. APT Bulletin the Journal of Preservation Technology, 2017, 48(1):6-14

[7] Brady, H. E. (2008). Causation and explanation in social science. The Oxford Handbook of Political Methodology.

[8] Bulleit, W., Schmidt, J., Alvi, I., Nelson, E., & Rodriguez-Nikl, T. (2015). Philosophy of engineering: What it is and why it matters. Journal of Professional Issues in Engineering Education and Practice, 141(3), 02514003.

[9] BungeM. Causality and modern science, 2017 Routledge

[10] BzdokD, AltmanN, KrzywinskiM. Statistics versus machine learning. Nature Methods, 2018, 15(4):233-234

[11] Chambliss, D. F., and Schutt, R. K. (2013). “Causation and experimental design.” Making sense of the social world: Methods of investigation.

[12] Chern, J.-C., You, C.-M., and Bazant, Z. P. (1992). “Deformation of Progressively Cracking Partially Prestressed Concrete Beams.” PCI Journal, 37(1), 74–85.

[13] ChristodoulouE, MaJ, CollinsGS, SteyerbergEW, VerbakelJY, Van CalsterB. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models. Journal of Clinical Epidemiology, 2019, 110: 12-22

[14] DegtyarevVV. Neural networks for predicting shear strength of CFS channels with slotted webs. Journal of Constructional Steel Research, 2021, 177: 106443

[15] Fernández-DelgadoM, SirsatMS, CernadasE, AlawadiS, BarroS, Febrero-BandeM. An extensive experimental survey of regression methods. Neural Networks, 2019, 111: 11-34

[16] FreundY, SchapireRE. A decision-theoretic generalization of on-line learning and an application to boosting. Journal of Computer and System Sciences, 1997, 55: 119-139

[17] Friedman, J. H. (2001). Greedy function approximation: A gradient boosting machine. Annals of Statistics, 29(5), 1189–1232.

[18] GoldbergDE, HollandJH. Genetic algorithms and machine learning. Machine Learning., 1988, 3: 95-99

[19] Gomez-Rubio, V. (2018). Generalized additive models: An introduction with R (2nd Edition). Journal of Statistical Software.

[20] GrayHJ. Empiricism in engineering and science. Science, 1965, 147: 557-558

[21] HandDJ. Probability for statistics and machine learning: fundamentals and advanced topics by anirban dasgupta. International Statistical Review, 2013, 81: 155

[22] HollandPW. Statistics and causal inference. Journal of the American Statistical Association, 1986, 81: 945-960

[23] ImbensGW. Potential outcome and directed acyclic graph approaches to causality: Relevance for empirical practice in economics. Journal of Economic Literature, 2020, 58(4):1129-1179

[24] Imbens, G. W., and Rubin, D. B. (2015). Causal inference: For statistics, social, and biomedical sciences an introduction. Causal Inference: For Statistics, Social, and Biomedical Sciences an Introduction.

[25] Keras. (2020). GitHub - keras-team/keras: Deep Learning for humans. https://github.com/keras-team/keras (Feb. 9, 2021).

[26] LeuridanB, WeberE. Causality and explanation in the sciences. Theoria (spain), 2012, 27(2):133-136

[27] Li, H., Xu, Z., Taylor, G., Studer, C., & Goldstein, T. (2018). Visualizing the loss landscape of neural nets. Advances in Neural Information Processing Systems, 31, 6389–6399.

[28] Lundberg, S. M., Erion, G. G., & Lee, S. I. (2018). Consistent individualized feature attribution for tree ensembles. Preprint at https://arxiv.org/abs/1802.03888v3.

[29] Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 4768–4777.

[30] MariniMM, SingerB. Causality in the social sciences. Sociological Methodology, 1988, 18: 347-409

[31] Molnar, C. (2019). Interpretable machine learning. A guide for making black box models explainable. https://christophm.github.io/interpretable-ml-book. Accessed 6 Jun 2018.

[32] MousaviSM, AminianP, GandomiAH, AlaviAH, BolandiH. A new predictive model for compressive strength of HPC using gene expression programming. Advances in Engineering Software, 2012, 45(1):105-114

[33] NaserMZ. Mechanistically informed machine learning and artificial intelligence in fire engineering and sciences, 2021 Springer

[34] NaserMZ. An engineer’s guide to eXplainable Artificial Intelligence and Interpretable Machine Learning: Navigating causality, forced goodness, and the false perception of inference. Automation in Construction, 2021, 129

[35] NaserMZ. Mapping functions: A physics-guided, data-driven and algorithm-agnostic machine learning approach to discover causal and descriptive expressions of engineering phenomena. Measurement, 2021, 185

[36] Naser, M. Z. (2022). Causality, causal discovery, and causal inference in structural engineering. https://doi.org/10.48550/arXiv.2204.01543.

[37] NaserMZ, AlaviAH. Error metrics and performance fitness indicators for artificial intelligence and machine learning in engineering and sciences. Architecture, Structures and Construction, 2021, 1: 1-19

[38] Naser, M. Z., & Ciftcioglu, A. O. (2022). Causal discovery and causal learning for fire resistance evaluation: incorporating domain knowledge. https://doi.org/10.48550/arXiv.2204.05311.

[39] NaserMZ, KodurV, ThaiH-T, HawilehR, AbdallaJ, DegtyarevVV. StructuresNet and FireNet: Benchmarking databases and machine learning algorithms in structural and fire engineering domains. Journal of Building Engineering, 2021, 44: 102977

[40] Pearl, J. (2009). Causal inference in statistics: An overview. Statistics Surveys.

[41] Pearl, J., Mackenzie, D. (2018). The Book of Why: The New Science of Cause and Effect-Basic Books. Notices of the American Mathematical Society.

[42] PearsonK. The grammar of science. Nature, 1892, 46: 247

[43] RandolphMF. Science and empiricism in pile foundation design. Geotechnique, 2003, 53(10):847-875

[44] Robins M. James, M. A. H. (2020). “Causal inference—what if.” Foundations of Agnostic Statistics.

[45] RoyPP, RoyK. On some aspects of variable selection for partial least squares regression models. QSAR and Combinatorial Science, 2008, 27(3):302-313

[46] RubinDB. Causal inference using potential outcomes. Journal of the American Statistical Association, 2005, 100: 322-331

[47] RudinC. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 2019, 1(5):206-215

[48] SadeghianV, VecchioF. The modified compression field theory: Then and now, 2018 ACI Special Publication

[49] Scikit. (2021). “sklearn.ensemble.AdaBoostRegressor—scikit-learn 0.24.1 documentation.” https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.AdaBoostRegressor.html (Mar. 10, 2021).

[50] Sivanandam, S. N., & Deepa, S. N. (2008). Genetic algorithm optimization problems. Introduction to Genetic Algorithms, Springer.

[51] SmithG. Probability and statistics in civil engineering, 1986 Collins

[52] SolhmirzaeiR, SalehiH, KodurV, NaserMZ. Machine learning framework for predicting failure mode and shear capacity of ultra high performance concrete beams. Engineering Structures, 2020, 224: 111221

[53] SunH, BurtonHV, HuangH. Machine learning applications for building structural design and performance assessment: State-of-the-art review. Journal of Building Engineering, 2021, 33

[54] VecchioFJ, CollinsMP. Modified compression-field theory for reinforced concrete elements subjected to shear. Journal of the American Concrete Institute, 1986, 83: 219-231

[55] XGBoost Python Package. (2020). “Python Package Introduction—xgboost 1.4.0-SNAPSHOT documentation.” https://xgboost.readthedocs.io/en/latest/python/python_intro.html#early-stopping (Feb. 10, 2021)

[56] XiongZ, CuiY, LiuZ, ZhaoY, HuM, HuJ. Evaluating explorative prediction power of machine learning algorithms for materials discovery using k-fold forward cross-validation. Computational Materials Science, 2020, 171: 109203

[57] XuL, LinSY, HlynkaAW, LuH, KamatVR, MenassaCC, El-TawilS, PrakashA, SpenceSMJ, McCormickJ. Distributed simulation platforms and data passing tools for natural hazards engineering: Reviews, limitations, and recommendations. International Journal of Disaster Risk Science, 2021, 12(5):617-634

[58] ZhangY, BurtonHV, SunH, ShokrabadiM. A machine learning framework for assessing post-earthquake structural safety. Structural Safety., 2018, 72: 1

[59] ZiegelER. The elements of statistical learning. Technometrics, 2003, 45(3):267-268